LUND UNIVERSITY

PO Box 117221 00 Lund+46 46-222 00 00

The Reproducibility of Berg Balance Scale and the Single-Leg Stance in ChronicStroke and the Relationship Between the Two Tests.

Flansbjer, Ulla-Britt; Blom, Johanna; Brogårdh, Christina

Published in:PM&R

DOI:10.1016/j.pmrj.2011.11.004

2012

Link to publication

Citation for published version (APA):Flansbjer, U-B., Blom, J., & Brogårdh, C. (2012). The Reproducibility of Berg Balance Scale and the Single-LegStance in Chronic Stroke and the Relationship Between the Two Tests. PM&R, 4(3), 165-170.https://doi.org/10.1016/j.pmrj.2011.11.004

Total number of authors:3

General rightsUnless other specific re-use rights are stated the following general rights apply:Copyright and moral rights for the publications made accessible in the public portal are retained by the authorsand/or other copyright owners and it is a condition of accessing publications that users recognise and abide by thelegal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private studyor research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal

Read more about Creative commons licenses: https://creativecommons.org/licenses/Take down policyIf you believe that this document breaches copyright please contact us providing details, and we will removeaccess to the work immediately and investigate your claim.

The Reproducibility of Berg Balance Scale and the Single-Leg

Stance in Chronic Stroke and the Relationship between the Two

Tests

Ulla-Britt Flansbjer, RPT, PhD1,2, Johanna Blom RPT3 and Christina Brogårdh, RPT, PhD1,2.

1Department of Rehabilitation Medicine, Skåne University Hospital, Lund, Sweden, 2Department of Health Sciences, Lund University, Lund, Sweden and 3Department of

Neurology, Skåne University Hospital, Malmö, Sweden

Short title: Reproducibility of balance tests in chronic stroke

Corresponding author: Ulla-Britt Flansbjer, RPT, PhD, Department of Rehabilitation

Medicine, Skåne University Hospital, SE-221 85 Lund, Sweden.

E-mail: ulla-britt.flansbjer@med.lu.se

The study was completed within the context of the Centre for Ageing and Supportive

Environments (CASE) at Lund University, funded by the Swedish Council for Working Life

and Social Research. Financial support was also received from the Swedish Stroke

Association, Magn. Bergvall Foundation, the Swedish Association of Neurologically Disabled

(NHR), the Swedish Society of Medicine, Gun and Bertil Stohne Foundation, the Crafoord

Foundation, the Council for Medical Health Care Research in South Sweden, Lund University

Hospital and Region Skåne.

ABSTRACT

Objective: To assess the reproducibility of the Berg Balance Scale (BBS) and the Single-leg

stance (SLS) and the validity of the SLS as an independent test of upright postural control in

patients with chronic stroke

Design: An intra-rater test-retest reproducibility study. The BBS and the SLS were assessed

twice, 7 days apart.

Setting: A university hospital

Participants: Fifty individuals; 6-46 months post-stroke.

Intervention: Not applicable

Main Outcome Measurements: The reproducibility of the BBS and the SLS was evaluated

with intraclass correlation coefficient (ICC2,1), the mean difference between the 2 test sessions

(d ̄ ) with 95% confidence interval (95% CI), the standard error of measurement (the standard

error of measurement [SEM]%), the smallest real difference (SRD%) and the Bland-Altman

graphs. To assess validity of SLS, the relationship between the SLS and the BBS was

analyzed by the Pearson correlation coefficient.

Results: The ICC2,1 was 0.88 for the BBS and the ICC2,1 values were 0.88 for the nonparetic

limb and 0.92 for the paretic lower limb for the SLS. The smallest change that indicates a real

improvement for a group of individuals, SEM%, was 3% for BBS, 15% for the nonparetic

limb and 27% for the paretic limb for SLS. The smallest real difference for a single individual

was 8% for BBS but was higher for SLS, at 42% for the nonparetic limb and 74% for the

paretic limb. There was a significant relationships between the SLS and the BBS (r= 0.65-

0.79; p< .001).

Conclusions: The BBS and the SLS are reproducible measurements in patients with chronic

stroke, but only the BBS is sensitive enough to follow changes over time or after an

intervention. The SLS is strongly related to the BBS and can be used as an independent test to

measure upright postural control after a stroke.

Key words: Cerebrovascular Disorders; Postural Balance; Outcome Assessment; Validity of

Results; Reproducibility of Results.

INTRODUCTION

Impaired balance is a common symptom after a stroke. Balance or postural control can be

defined as the ability to maintain, achieve, or restore stability of posture or during activity [1].

Balance deficits after a stroke can be caused by lesions in the pons or the cerebellum, or as a

consequence of hemiparesis, sensory impairment, reduced visual field, or cognitive deficits

[1-3]. Individuals after a stroke have shown 4-5 times greater postural instability in both the

frontal and sagittal planes compared with healthy age-matched individuals [3]. Reduced

balance after stroke is associated with low ambulatory activity [4] and increased risk of falling

[5].

Many assessment tools are used to evaluate balance, for example, the Berg Balance

Scale (BBS), Functional Reach, Postural Assessment Scale for Stroke patients, Single-leg

Stance (SLS) and Computerized Dynamic Posturography [2, 6-8]. The most common

instrument to measure balance after stroke is the BBS, which has been shown to be a valid

and reliable test of upright postural control in a variety of populations, including patients who

have experienced a stroke [9-11]. The sensitivity to change has been assessed in the acute

and/or subacute phase after a stroke [12] but is lacking in the chronic phase.

Another common balance test is the SLS (also called “one-leg standing”), frequently

used for elderly populations. The test-retest reliability has been shown to be acceptable [13-

15], even if the testing procedures vary in different studies [16]. The SLS is less time

consuming than the BBS, which can be of importance for clinicians [6]. One-leg standing is

included as one item in the BBS, and has been shown to have a good inter-item and item-to-

total correlation in patients with an acute stroke [9]. However, the validity and reliability of

the SLS as an independent test of upright postural control has not been assessed in patients

who have experienced a stroke.

To be able to evaluate balance after a stroke, reliable and sensitive testing methods are

essential. Reproducibility in clinical practice and medical research can be determined from

measurements of the same individuals on 2 occasions, so-called test-retest reproducibility. To

fully assess the reproducibility, several statistical methods are required, which cover both

agreements between measurements, systematic changes in the mean, and measurement errors

[17]. In addition, a measurement tool can be considered highly reliable, but may not be

sufficiently sensitive to detect a real change over time or after an intervention.

The aim of this study was to evaluate (1)) the reproducibility of the BBS and SLS as

balance tests in patients with chronic stroke and to define limits for the smallest change that

indicates a real improvement, both for a group of individuals and for a single individual and

(2) the validity of the SLS as an independent test of upright postural control after a stroke by

analyzing the relationship between the SLS and the BBS.

MATERIAL AND METHODS

Individuals

All participants were recruited from a database at a rehabilitation unit in a university hospital

in the south of Sweden. Fifty community-dwelling individuals (38 men and 12 women; mean

± standard deviation [SD] age, 58 ± 6 years; range, 46-72 years; mean [± SD] time since

stroke onset, 17 ± 9 months; range, 6-46 months) met the following inclusion criteria: (1) a

minimum of 6 months post-stroke; (2) residual hemiparesis at discharge from primary

rehabilitation; (3) able to understand both verbal and written information; (4) able to walk at

least 300 meters with or without a unilateral assistive device; (5) able to stand without hand-

held support; and (6) medically stable, with no other diseases that could significantly

influenced muscle strength, gait performance or postural control. The clinical characteristics

of the individuals are presented in Table 1. Before the first test session, all the individuals

completed a questionnaire, which provided demographic and medical information. All

individuals were checked by the responsible physician to fulfill the inclusion criteria and to be

medically stable and therefore suitable to participate in the study. At the time of the

assessments the individuals also participated in 2 other reliability studies of gait performance

tests and knee muscles strength measurements [18, 19].

Insert Table 1 about here

Ethics Considerations

All the individuals were contacted by telephone, received verbal and written information

about the study, and then gave their informed consent. The ethics research committee of Lund

University, Lund, Sweden approved the study (LU 243-01).

Balance Tests

Each participant underwent the BBS and the SLS on 2 occasions. The BBS was tested before

SLS at both test occasions. The BBS was performed by following the standard procedure

[10]. The test is a 14-item scale, scored from 0 to 4; a score of 0 represents the inability to

complete the task and a score of 4 represents independent item completion. The sum score of

all 14 items (maximum score for the BBS, 56 points) was calculated and used in the analyses.

When the SLS was performed, the participants were instructed to stand on one leg, as long as

possible, but not longer than 15 seconds, starting with the preferred leg. The SLS was timed

when one foot was lifted from the floor until it touched the ground or the other leg; no hand-

held support was allowed during the test. The SLS was repeated 3 times during each session,

with the right and left foot alternately, and the mean times for the 3 trials were then

determined. To avoid any learning effect the first trial of SLS was used to score the last item

in BBS; one-leg standing on the preferred leg in the BBS is scored from 0 point (not able to

lift one foot or need help not to fall) to 4 points (if able to stand on one leg at least 10

seconds).

Procedure

The individuals were tested in a rehabilitation unit in a university hospital on 2 occasions, 7

days apart, at the same time of the day. The participants were instructed not to change their

normal physical activities between the 2 test occasions. One senior physiotherapist (U.-B.F.)

did all assessments and has extensive experience from stroke rehabilitation and the tests used.

The test protocols for BBS and SLS were carefully standardized. No verbal encouragement

was given during the tests. Throughout each session, the individuals wore comfortable shoes

and were allowed to use their common ankle-foot orthosis (n=7) but no other assistive

devices. The total time for the 2 balance tests was approximately 45 minutes. After the first

test session, the individuals received information about the second test session but were not

informed about their results. A written summary and oral information about the test results

were given after completion of the second test session. All individuals were provided

transportation free of charge to and from the test site. The test procedure (ie, time interval

between the test occasions, standardized protocols, verbal information during the tests) has

previously been developed and used in 2 other studies in our research group [18, 19].

Data and statistical analyses

All 50 individuals completed the tests. The recorded variables, obtained from the 2 test

sessions, were used in the analysis. Descriptive statistics (means and SD) were calculated for

the BBS and the SLS. To determine the test-retest reproducibility for the BBS and the SLS,

several statistical methods were applied [17, 20]. Agreement between measurements was

analyzed by the intraclass correlation coefficients, (ICC2,1) and the mean differences between

the test sessions (d ̄ ) together with the 95% confidence intervals (CI) for d ̄ [21]. Measurement

errors were assessed by the standard error of measurement (SEM) and the SEM%. The SEM

gives the measurement errors in absolute values and represents the limit for the smallest

change that indicates a real change for a group of individuals. The SEM% is independent of

the units of measurement and therefore more easily interpreted. The smallest real difference,

(SRD) [22], which represents the limits for the smallest change that indicates a real change for

a single individual, was calculated, together with an “error band” around the mean difference

of the 2 measurements, d ̄ . From the SRD, the SRD% was calculated, which represents the

change in relative terms.

The Bland-Altman graphs were formed to give a visual interpretation of the data [23].

From each test session, the relationship between the SLS (the mean of the 3 trials from each

lower limb) and the sum score of BBS was analyzed using the Pearson product moment

correlation coefficient. All calculations were performed using the SPSS 18.0 Software for

Windows (SPSS Inc, Chicago, IL). A significance level greater than .05 represented

nonsignificance.

RESULTS

None of the participants reported any negative events between or during the 2 test sessions,

which potentially could have influenced the results. The mean values from the 2 test sessions

for the BBS and the SLS are presented in Table 2.

Insert Table 2 about here

Reproducibility

The reproducibility of the BBS and the SLS are presented in Table 3. The ICC2,1 values were

0.88 for BBS and were 0.88 for the nonparetic lower limb and 0.92 for the paretic lower limb

for the SLS. The d̄ values were close to zero, and the widths of the 95% CI for d̄ were narrow,

which demonstrated a small distribution. A positive value of d̄ was found for the BBS and for

the SLS nonparetic lower limb, which means that the performance at the second test session

was better than at the first (p< .05), which in turn indicates a learning effect. The SEM% was

low for BBS (3%) but higher for the SLS (15% for the nonparetic and 27% for the paretic

limb). The SRD% was also low for the BBS (8%) but higher for SLS (42% for the nonparetic

limb and 74% for the paretic limb). From the Bland-Altman graphs (Figure 1), a systematic

variation around the zero line was revealed. In general, there were more values above the zero

line than below for the BBS and for the SLS nonparetic lower limb, which illustrates the

better performance at the second test session.

Insert Table 3 and Figure 1 about here

Validity

The relationship between the SLS and the BBS was determined by data from the 2 test

sessions. The correlations between the SLS and the BBS were significant (p< .001) for both

lower limbs at both test sessions. The r value for the nonparetic lower limb was .77 at the first

occasion and .79 at the second occasion, and for the paretic lower limb .7 at the first occasion

and .65 at the second occasion.

DISCUSSION

The main finding of this study is that although both the BBS and the SLS are reproducible

measurements in patients with chronic stroke, only the BBS is sensitive enough to follow

changes over time or after an intervention. Because there are strong relationships between the

SLS and the BBS, the results indicate that the SLS could be used as an independent test of

upright postural control in chronic stroke.

Reliable and sensible measurement tools are essential to be able to follow changes over

time or after an intervention. To fully investigate the reproducibility of a test, several

statistical methods were applied in this study. Hopkins [24] suggested that a sample size of at

least 30 individuals should be considered in reliability studies, but a larger sample size gives

safer results for a given population [25], and therefore a total of 50 individuals were recruited

in this study. Furthermore, much attention was paid to the procedure of the test protocol. All

conditions were as stable as possible: for example the same assessor, assessments at the same

time of the day, and standardized instructions during the tests to reduces the possibility of

errors. The tests were performed at the same weekday and at the same time of the day, and all

individuals were told to live as usual in the meantime. The design used in this study was

previously developed and used within our research group. For patients in a chronic phase after

a stroke, no change in balance deficits would be expected within 1 week. Because no

participants reported any negative events between or during the 2 test sessions and the

protocol was strictly standardized, the change in balance could be considered as normal

variations within the individuals.

The ICC2,1 was chosen because it provides the basis for the calculations of the SEM. By

following the guidelines of Shrout and Fleiss [21], an ICC1,1 might, be the most correct

equation. However, in practice, the values of the different ICCs are often more or less

identical and it has been suggested that an ICC2,1 can be used in most circumstances. The

results from this study showed that the ICC2,1 were high for both the BBS (0.88) and the SLS

(0.88 and 0.92 for the non-paretic and paretic lower limb, respectively) with low measurement

errors. Our results for BBS are comparable with the results from another reliability study of

BBS in the early phase after a stroke with an ICC2.1 of 0.92 for individuals able to stand

without any support [12].

Even if the ICCs were high for both measurements, the variance differed, with low

values for the BBS but higher values for the SLS, which yielded higher values for the SEM%

and the SRD%, especially for the paretic leg. From a clinical point of view, only the SRD%

values for the BBS (8%) were sufficiently small to detect real changes in balance for single

individuals with chronic stroke; for example, for an individual with a sum score of 50 points

for the BBS, a change of 4 points is needed to detect a real improvement. For an individual

able to stand for 7 seconds on one leg, a real change in the SLS, would be at least 3 seconds

for the non-paretic lower limb and 5 seconds for the paretic lower limb, which clearly

illustrates the need for several statistical methods to fully assess the reliability of a

measurement method.

The BBS was assessed before the SLS at both test session. Because one-leg standing is

one item in the BBS, there could have been a learning effect from the BBS to the SLS. To

avoid this learning effect, the last item of BBS was scored from the first trial of SLS which

could be done because the instructions for the last item (one-leg standing) in BBS was similar

to those used in the SLS and the time frame was longer for SLS (15 seconds compared to 10

seconds for BBS). A better performance during the second test session was found for both the

BBS and the non-paretic lower limb of the SLS, which suggests a small learning effect, which

also has been reported in gait performance after stroke [17]. Even though this learning effect

was small, it has to be taken into account. Learning effects might be reduced if practice is

allowed before the actual test procedure. However, when following the standardized protocol

of the BBS, a practice before the assessment is not allowed. To reduce a possible learning

effect of SLS, the participants were allowed to perform 3 trials and the mean values were used

in the statistical analysis. Still, we found a better performance for SLS at the second test

session than the first, but only for the nonparetic lower limb.

Limitations

In this population, the balance varied within the group, and these differences were also

demonstrated by the need of assistive device for 18 of the 50 participants. All participants

were well recovered, able to stand without support and to walk with or without an assistive

device; however the result from this study are representative only for persons with mild-to-

moderate disability after a stroke.

A problem with the SLS is that it has not been fully standardized. Leg selection,

maximum times, opened or closed eyes, restrictions about the arms, and the number of trials

vary [13, 16]. The design of SLS in this study was standardized before the assessments, with a

maximum time of 15 seconds; the preferred leg was assessed first, with eyes opened; and

aspects of security were highlighted so as not to cause any risk for the participants. In further

studies the optimal standardization of the SLS in patients with chronic stroke should be

addressed.

The results from this study could guide the clinicians to choose appropriate

measurements when assessing balance in patients with chronic stroke. The BBS could be

considered as the criterion standard test to measure balance in stroke; however the SLS is a

less time-consuming balance test, does not require any special equipment or training and is

performed on both the paretic and nonparetic lower limbs. Both assessments can be used in

clinical settings but only BBS seems to be appropriate for follow-up evaluations after an

intervention or over time.

CONCLUSIONS

The BBS and the SLS are reproducible measurement tools after stroke, but the SLS is less

sensitive for measuring changes over time or after an intervention. There are strong

relationships between the BBS and the SLS for both lower limbs which indicates that the SLS

is a valid test of upright postural control for both lower limbs, in patients after a stroke and can

be recommended as a test of balance in patients with chronic stroke.

REFERENCES

1. Pollock AS, Durward BR, Rowe PJ, Paul JP: What is balance? Clin Rehabil

2000;14:402-406.

2. de Oliveira CB, de Medeiros IR, Frota NA, Greters ME, Conforto AB: Balance control

in hemiparetic stroke patients: main tools for evaluation. J Rehabil Res Dev

2008;45:1215-1226.

3. de Haart M, Geurts AC, Huidekoper SC, Fasotti L, van Limbeek J: Recovery of

standing balance in postacute stroke patients: a rehabilitation cohort study. Arch Phys

Med Rehabil 2004;85:886-895.

4. Michael KM, Allen JK, Macko RF: Reduced ambulatory activity after stroke: the role

of balance, gait, and cardiovascular fitness. Arch Phys Med Rehabil 2005;86:1552-

1556.

5. Lamb SE, Ferrucci L, Volapto S, Fried LP, Guralnik JM: Risk factors for falling in

home-dwelling older women with stroke: the Women's Health and Aging Study. Stroke

2003;34:494-501.

6. Tyson SF, Connell LA: How to measure balance in clinical practice. A systematic

review of the psychometrics and clinical utility of measures of balance activity for

neurological conditions. Clin Rehabil 2009;23:824-840.

7. Lubetzky-Vilnai A, Kartin D: The effect of balance training on balance performance in

individuals poststroke: a systematic review. J Neurol Phys Ther 2010;34:127-137.

8. Orr R, Raymond J, Fiatarone Singh M: Efficacy of progressive resistance training on

balance performance in older adults : a systematic review of randomized controlled

trials. Sports Med 2008;38:317-343.

9. Berg K, Wood-Dauphinee S, Williams JI: The Balance Scale: reliability assessment

with elderly residents and patients with an acute stroke. Scand J Rehabil Med

1995;27:27-36.

10. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B: Measuring balance in the elderly:

validation of an instrument. Can J Public Health 1992;83 Suppl 2:S7-11.

11. Blum L, Korner-Bitensky N: Usefulness of the Berg Balance Scale in stroke

rehabilitation: a systematic review. Phys Ther 2008;88:559-566.

12. Stevenson TJ, Garland SJ: Standing balance during internally produced perturbations in

subjects with hemiplegia: validation of the balance scale. Arch Phys Med Rehabil

1996;77:656-662.

13. Michikawa T, Nishiwaki Y, Takebayashi T, Toyama Y: One-leg standing test for

elderly populations. J Orthop Sci 2009;14:675-685.

14. Franchignoni F, Tesio L, Martino MT, Ricupero C: Reliability of four simple,

quantitative tests of balance and mobility in healthy elderly females. Aging (Milano)

1998;10:26-31.

15. Giorgetti MM, Harris BA, Jette A: Reliability of clinical balance outcome measures in

the elderly. Physiother Res Int 1998;3:274-283.

16. Bohannon RW: Single Limb Stance Times: A Descriptive Meta-Analysis of Data From

Individuals at Least 60 Years of Age. Top Geriatr Rehabil 2006;22:70-77.

17. Lexell JE, Downham DY: How to Assess the Reliability of Measurements in

Rehabilitation. Am J Phys Med Rehabil 2005;84:719-723.

18. Flansbjer U-B, Holmbäck AM, Downham D, Patten C, Lexell J: Reliability of gait

performance tests in men and women with hemiparesis after stroke. J Rehabil Med

2005;37:75-82.

19. Flansbjer U-B, Holmbäck AM, Downham D, Lexell J: What change in isokinetic knee

muscle strength can be detected in men and women after stroke? Clin Rehabil

2005;19:514-522.

20. Atkinson G, Nevill AM: Statistical methods for assessing measurement error

(reliability) in variables relevant to sports medicine. Sports Med 1998;26:217-238.

21. Shrout PE, Fleiss JL: Intraclass correlations: Uses in assessing rater reliability. Psychol

Bull 1979;86:420-428.

22. Beckerman H, Roebroeck ME, Lankhorst GJ, Becher JG, Bezemer PD, Verbeek AL:

Smallest real difference, a link between reproducibility and responsiveness. Qual Life

Res 2001;10:571-578.

23. Bland JM, Altman DG: Statistical methods for assessing agreement between two

methods of clinical measurement. Lancet 1986;1:307-310.

24. Hopkins WG: Measures of reliability in sports medicine and science. Sports Med

2000;30:1-15.

25. Donner A, Eliasziw M: Sample size requirements for reliability studies. Stat Med

1987;6:441-448.

LEGENDS

Figure 1

The differences between test sessions 2 and 1 (test 2 minus test 1) plotted against the means

of the two test sessions for the Berg Balance Scale (BBS) and the Single-leg Stance (SLS) for

the nonparetic and the paretic lower limb. From these Bland- Altman graphs, the systematic

variation around the zero line was revealed, and shows the learning effect for the BBS and

nonparetic lower limb in the SLS.

SLS paretic lower limb

0 3 6 9 12 15

Diff

eren

ce (

s)

-8

-4

0

4

8

SLS non-paretic lower limb

0 3 6 9 12 15

Diff

eren

ce (

s)

-8

-4

0

4

8

Mean (s) Mean (s)

Berg Balance Scale

Mean (points)

40 44 48 52 56D

iffer

ence

(po

ints

)-10

-5

0

5

10

Figure 1

Table 1. Clinical characteristics of the 50 individuals post stroke.

Characteristics n (%)

Type of stroke

Ischemic 37 (74)

Hemorrhagic 13 (26)

Hemiparetic side

Weakness in right side 20 (40)

Weakness in left side 30 (60)

Use of assistive device

No walking aid 32 (64)

Walking aid 11 (22)

Ankle-foot orthosis and walking aid 7 (14)

Table 2. Berg Balance Scale and Single-leg stance results for 50 individuals after

stroke.

Test

Test session 1

Mean ± SD

Test session 2

Mean ± SD

Berg Balance test, points 52 ± 4.3 52.7 ± 3.8

Single-leg stance, s

Nonparetic 10.6 ± 5 11.5 ± 4.7

Paretic 6.6 ± 6 6.5 ± 6.1

Table 3. Reproducibility of the Berg Balance Scale and the Single-leg stance for the 50 individuals after stroke.

Test ICC2,1 (95% CI) d̄ (95% CI) SEM SEM% SRD 95% SRD SRD%

Berg Balance Scale, points 0.88 (0.80-0.93) 0.72 (0.15-1.29) 1.49 3 4.13 -3.41 to 4.85 8

Single-leg stance, s

Nonparetic 0.88 (0.79-0.94) 0.82 (0.19-1.45) 1.66 15 4.59 -3.77 to 5.41 42

Paretic 0.92 (0.86-0.95) -0.03 (-0.74-0.68) 1.76 27 4.87 -4.90 to 4.84 74

ICC2.1 = intraclass correlation coefficient; CI = confidence interval; d ̄ = the mean differences between the test sessions; SEM =

standard error of measurement; SRD = smallest real difference